Tze Leung
Papers
1
Total Citations
3
H-Index
1
About
Tze Leung is a leading figure in statistical signal processing and adaptive control, renowned for pioneering work on fast particle filters and change-point detection. His research centers on developing computationally efficient algorithms for systems that undergo sudden, unobserved parameter shifts—a challenge critical to robotics, autonomous navigation, and real-time control. Leung’s major contribution lies in bridging Bayesian inference with online learning: he introduced the AFMM (adaptive forgetting through multiple models) framework, which dynamically updates posterior probabilities across a family of models to robustly track abrupt changes without requiring manual resetting. His 2009 paper on fast particle filters, though accruing modest citations, laid foundational theory for real-time adaptation in change-point ARX models and robotic systems, demonstrating how carefully designed change detection can segment data streams on the fly. Beyond this, Leung’s work has influenced adaptive control strategies in environments where sensor noise and system dynamics evolve unpredictably. His achievements include advancing the theoretical underpinnings of sequential Monte Carlo methods for non-stationary processes, earning recognition among engineers and statisticians for making complex Bayesian updates practical in latency-sensitive applications. For students, Leung’s research exemplifies how rigorous statistical theory can drive tangible innovations in autonomous systems and adaptive robotics.
Research Focus
Key Achievements
Top Papers
- 1